Guide
Columns called x193iq5w are not your data's fault.
You export a list and the headers read x193iq5w, x1i10hfl href, css-1x2y3z. Some cells say [object Object]. Half the columns are empty, and tags are spread over Tag 2 to Tag 14. It is a common sight with one-click scrapers, and it comes from how they name what they find.
Where names like x193iq5w come from
The column is named after a CSS class
A scraper that does not know what a value is names its column after the element's class. Modern sites generate those classes when they are built: x193iq5w, css-1x2y3z, sc-bdVaJa. They mean nothing to a person and change whenever the site is rebuilt.
[object Object] is a value that was never turned into text
A tool that copies a structure where it should copy its text writes the programming language's placeholder into the cell. The data was there; the tool did not read it out.
Tags become Tag 2 to Tag 14
A record with a varying number of the same thing — tags, categories, thumbnails — comes out as one column per position, each mostly empty.
The same link appears five times
A card wrapped in a link, with the title, picture and button all linking to the same page, gives the same address in several columns.
How Magic Scraper names its columns
Readable names, without setting anything up.
From the page's own headers and labels first
A table's header row, two-row headers included; a definition list's terms; "Phone:" and "Address:" labels beside their values; the page's structured data.
Then from what the values are
A column of prices is Price, dates are Date, addresses Link, pictures Image, and a class name is used only when it is a real word like product-title, never a build hash.
Values are always text
Every cell is the text you see, the address a link goes to, or the picture's address. There is no [object Object].
Repeats are tidied
A column identical to another is dropped, an image's alt text that repeats the name is folded in, and tags of varying number are merged into one Tag column next update.
Still not right? Click a column header to drop it, double-click it to rename it, and the names you choose are remembered for that site and used in every export.
Side by side
| Some one-click scrapers | Magic Scraper | |
|---|---|---|
| Column names | CSS classes: x193iq5w, x1i10hfl href | Price, Date, Link, Image, the page's own headers |
| Objects in cells | [object Object] | Always the visible text, or the link or image address |
| Tags and categories | Tag 2 … Tag 14, mostly empty | One Tag column next update |
| Rename and drop | Varies | Click to drop, double-click to rename, remembered per site |
Reasonable questions
Why are my scraped column names random letters?
They are the page's CSS class names, which modern sites generate automatically. A scraper that falls back on them gives you names like x193iq5w. Magic Scraper names columns from the page's headers and labels, or from the values, instead.
What does [object Object] mean in a scraped spreadsheet?
The tool copied a programming structure where it should have copied its text. The value was on the page but never read out. Magic Scraper writes the text you see, or the link or image address.
Can I rename columns?
Yes. Double-click a column header in the preview to rename it, click it to drop it. Your choices are remembered for that site.
Why do I get Tag 2, Tag 3 ... Tag 14?
Each record has a different number of tags and the tool made a column per position. The next update of Magic Scraper merges them into one Tag column, the tags separated by commas.